双种群教与学算法在磁轴承PID控制器中的应用

Application of Double Populations Teaching-Learning-Based Optimization in PID Controller of Magnetic Bearings

针对主动磁轴承系统的开环不稳定性,提出采用改进双种群教与学算法与PID结合的方法实现对系统的闭环控制。该算法在教与学算法对PID参数的优化中加入教师反向学习和学生变异学习,解决基本算法的早熟问题,提高解的精度。运用MATLAB/Simulink搭建系统模型进行仿真试验,结果表明:与基本教与学算法和遗传算法相比,该算法收敛到最优参数值的速度更快,PID控制磁轴承系统的响应速度更快,具有更好的动态性能和稳态性能。

The open-loop instability of active magnetic bearing system is studied,and the method of improved double populations teaching-learning-based optimization combined with PID is proposed to realize closed-loop control of the system. The algorithm integrates teacher backward learning and student variation learning during optimization of PID parameters in teaching-learning optimization,which solve premature problem about basic algorithm and improve precision of solution. The system model is built to simulate experiment in MATLAB/Simulink. The results show that compared with basic teaching-learning-based optimization and genetic algorithm,the algorithm faster converge to optimal parameter values,the PID controller faster response and has better dynamic and steady-state performances.